Regardless, no one should be committing to core without running the integration
tests.
> On May 24, 2016, at 2:28 AM, Hervé BOUTEMY wrote:
>
> notice there was an outage last week, causing lots of false positives for a
> few days, during the week-end
>
> but in general, in the last year, our
notice there was an outage last week, causing lots of false positives for a
few days, during the week-end
but in general, in the last year, our CI server is working as expected, with
few false positives (even if not perfect)
See
https://builds.apache.org/view/M-R/view/Maven/
I worked out some l
The upgrade of SLF4J was causing this. I did not expect a patch upgrade
of SLF4J from 1.7.16 to 1.7.21 to produce such an issue and hadn't run
the ITs locally. Somehow I got used to Jenkins sending "build failed"
mails after every commit for a reason not related to the commit. The
"build still
Am 2016-05-20 um 00:02 schrieb Anton Tanasenko:
Looks like the high failure count was due to
https://issues.apache.org/jira/browse/MNG-6023
The commit that reverted the slf4j version change came in some 9 hours ago
while we were running tests just before that.
https://github.com/apache/maven/comm
Looks like the high failure count was due to
https://issues.apache.org/jira/browse/MNG-6023
The commit that reverted the slf4j version change came in some 9 hours ago
while we were running tests just before that.
https://github.com/apache/maven/commit/4b35102f0e3302e15a80d3ddb756463601b9217a
2016-
Am 2016-05-19 um 13:29 schrieb Jason van Zyl:
Igor, Anton, and myself between OS X, Linux and Windows have 40 IT
failures in Maven master right now. I’ve had the same setup for 8
years, Igor and Anton have consistent setups as well. I’m not sure if
people aren’t running the integration tests or w
Hi Jason,
I tried the gist bash script.
First of all I’ve to change the patchUrl:
——8<—8<——snipp——8<—8<——
[…]
#patchUrl=${repository}/pull/${pr}.patch
patchUrl=https://patch-diff.githubusercontent.com/raw/apache/maven/pull/${pr}.patch
[…]
——8<—8<——snipp——8<—8<——
otherwise the cont
This is what I wrote for people trying to contribute to core:
http://takari.io/2014/06/01/contributing-to-maven-core.html
And what I use to smoke test contributed PRs:
https://gist.github.com/jvanzyl/16da25976f8ad27293fa
> On May 19, 2016, at 3:00 PM, Karl Heinz Marbaise wrote:
>
> Hi Jason,
Myself, Igor, Anton and this CI instance
(https://ci.takari.io/job/maven-integration-tests/246/console) get the same
thing.
So we have an even worse situation where some changes made to scripts or the IT
harness allow us all the get wildly different results.
> On May 19, 2016, at 3:20 PM, Uwe
Hi Jason,
I tried to reproduce the IT failure at master.
Fetched https://github.com/apache/maven - master - last commit:
4b35102f0e3302e15a80d3ddb756463601b9217a
Fetched https://github.com/apache/maven-integration-testing - master - last
commit: af138c8a75a0b1a852510b27ed2c469992cde69a
——8<———
Hi Jason,
i would like to know more details about that setup cause i'm thinking
for a longer time to make a setup of the IT environment via Docker etc.
this would mean having a consistent and reproducible environment for
running the it's...
Kind regards
Karl Heinz
On 5/19/16 1:29 PM, Jason v
Igor, Anton, and myself between OS X, Linux and Windows have 40 IT failures in
Maven master right now. I’ve had the same setup for 8 years, Igor and Anton
have consistent setups as well. I’m not sure if people aren’t running the
integration tests or what but it’s pretty sad. It’s not a small num
12 matches
Mail list logo
